Sample preparation for next-generation sequencing (NGS) is an inefficient, complex, multistep process that is difficult to automate. Sequencing library construction consists of around a dozen biochemical operations. Currently, automation of these steps requires highly customized automation equipment and highly trained technicians to manage the workflow. There are only a couple of automation products that address this problem, but these only perform part of the overall library construction process-they do not perform DNA purification from the biological sample. As NGS technology begins to move into the diagnostic world, there is an urgent need for walkaway, closed automated systems that dramatically simplify the NGS sample prep process. In addition, pioneering NGS companies and researchers are pushing the sequencing field toward very long-read, single-molecule sequencing technologies. This move to longer-read sequencing methods poses new challenges for NGS sample prep. Most popular DNA extraction methods do not produce long genomic DNA molecules, and many third-gen sequencing companies are actively seeking new, reliable methods for extracting large, intact DNA molecules, and converting them into sequencing libraries. In the course of developing new DNA size-selection technologies for NGS, we have conceived of a powerful technology for automated, rapid, purification of very long genomic DNA. In addition, the technology can be easily adapted to also produce NGS libraries for both traditional short-read and third-gen long-read sequencing platforms. By combining these two capabilities, our phase II application seeks to produce a system capable of producing finished NGS libraries from crude blood samples, in a single completely automated workflow. Such a system will dramatically reduce the cost and complexity of NGS sample prep, and speed the application of NGS to diagnostics.